CDC data indicate that 7% of the US population aged 55-64 experience limitation in activity due to arthritis, increasing to 19% for persons over age 84. When these data are projected to the year 2020, the number of persons with arthritis is estimated to increase by 57% and activity limitation by 66% (U.S., DHHS, 1994). Current data indicate that lower- extremity joint impairment is a serious risk factor for future disability. Previous exercise interventions have demonstrated short-term efficacy in improved interim performance measures (e.g., 6-minute distance walk) among older persons with lower-extremity osteoarthritis (OA). However, the long- term impact of a multiple component prevention intervention that blends exercise with education aimed at enhancing self-efficacy on maintenance of functional status is unknown. Using stages of behavioral change theory and social/cognitive learning theory, we hypothesize that a multiple component intervention that addresses lower-extremity muscle strength, fitness walking, and exercise and arthritis self-efficacy will delay disability onset. To test this hypothesis, we will convert a current NIAMS-funded randomized study of short-term efficacy (8 weeks) of a multiple component intervention to a longitudinal trial. The longitudinal trial will obtain adherence measures quality over 24 months on 146 treatments and 146 control group participants. Adherence is defined as 2 hours per week of self-reported exercise activity. To reinforce adherence, we will invite graduates of the short-term intervention to participate in an on-going once weekly facility-based walking program. The facility-based program will be supplemented with individualized regiments and supplies need for ongoing home-based adherence. We will use GEE to assess whether the treatment group maintains a significantly higher level of adherence over 24 months than controls, will test a multivariate model that predicts adherence, and will examine the impact of self-efficacy on sustained improvement in interim performance and adherence, and the impact of sustained improvement in interim performance on functional status outcomes over time.